KR20210037611A - Codon optimization - Google Patents

Abstract

Exemplary computer-implemented methods for optimizing nucleic acid sequences for protein expression in a host include: a) receiving an initial population set, wherein the initial population set is a plurality of initial populations capable of expressing a protein. Step 106, comprising a candidate nucleic acid sequence; And b) performing optimization of the harmonic index, codon context index and outlier index using a computer-assisted NSGA-III algorithm or a variant thereof based on the initial population set, whereby a plurality of proteins capable of expressing a protein Obtaining an optimized nucleic acid sequence (108).

Description

Codon optimization

Submitting a sequence listing in an ASCII text file

The contents of the following submissions on ASCII text files are incorporated herein by reference in their entirety: Computer-readable Format of Sequence Listing (CRF) (file name: 759892000440SEQLIST.TXT, date recorded: 25 July 2018, size: 4 KB).

Field of invention

The present disclosure relates generally to optimization techniques, and more particularly to systems and methods for optimizing sequences (eg, nucleic acid sequences) for protein expression in a host.

Codon weight refers to the redundancy of the genetic code as a phenomenon in which amino acids can be characterized by different synonym codons. In particular, it has been found that these synonym codons are used with unequal frequencies in most sequenced genomes. This phenomenon is called the codon-frequency bias.

Since biomedical and biotechnology research and industrial production require high-quality proteins with the correct folding and modifications, improving the expression levels of proteins is potentially reflecting the codon-use bias of highly-expressed genes. A way to explore and summarize useful rules and patterns is essential. However, protein expression is a multi-step process that involves regulation at the level of transcription, mRNA conversion, translation, and post-translational modifications that can form stable products. Even a single synonym codon substitution can increase the expression of a transgene by more than 1,000-fold. Thus, codon optimization is prepared for optimal expression of the synthetic gene in the recombinant host.

Brief overview

Systems and methods are provided herein for improved codon optimization that takes into account as well as balances multiple factors using a multi-goal optimization algorithm. According to some embodiments, codon optimization has, among other things, three goals: (i) how to initially allocate the coefficients of synonym codons for a particular amino acid, (ii) how to place synonym codons at the most suitable positions of the synonym codons, And (iii) a method of reducing unfavorable but accidentally generated subsequences and/or motifs. In some embodiments, these three goals are quantified as a harmonic index, a codon context index, and an outlier index. During optimization, targets are considered using a multi-goal algorithm such as Non-Dominant Classification Genetic Algorithm III (NSGA-III) or a variant thereof. Specifically, a target can be calculated for a given candidate nucleic acid sequence with reference to the known properties of the highly-expressed gene. In some embodiments, various known adverse motifs and/or features (eg, as identified in the literature) are removed from one or more optimized sequences prior to gene synthesis and protein expression.

Accordingly, the present invention is preferably thereby codon matching, codon usage (e.g., synonymous codon distribution), codon context index, cis-acting mRNA destabilization motif, RNase splicing site, GC-content, ribosome binding site. (RBS), the mRNA secondary structure of the gene (e.g., mRNA free energy) and all or most of the parameters and factors affecting protein expression including, but not limited to, repeating elements are mammals, insects, yeast, Considerations for improving and optimizing nucleic acid sequences to elevate protein expression of genes in expression host cells, including both eukaryotic and prokaryotic cells, such as bacteria and algae, and in expression systems, such as in cell-free expression systems. It provides a systematic way to be taken on.

In some embodiments, a) receiving an initial population set, wherein the initial population set comprises a plurality of initial candidate nucleic acid sequences capable of expressing a protein; And b) performing optimization of the harmonic index, codon context index and outlier index using a computer-assisted NSGA-III algorithm or a variant thereof based on the initial population set, whereby a plurality of proteins capable of expressing a protein Obtaining an optimized nucleic acid sequence, wherein the harmony index of the candidate nucleic acid sequence indicates the consistency of the frequency distribution of synonym codons between the plurality of highly expressed genes and the candidate nucleic acid sequence, wherein the codon context index of the candidate nucleic acid sequence Is a measure for placing synonym codons in appropriate positions, wherein the outlier index of the candidate nucleic acid sequence is a measure of the negative effect of a plurality of predetermined sequence features on the candidate nucleic acid sequence. Computer-implemented-methods for optimizing nucleic acid sequences for use are provided.

In some embodiments, the method further comprises providing an output representative of at least one optimized nucleic acid sequence of the plurality of optimized nucleic acid sequences.

In some embodiments, receiving the initial population set comprises receiving a protein sequence; And generating an initial population set based on the received protein sequence.

In some embodiments, receiving the initial population set comprises receiving a nucleic acid sequence; Translating the received nucleic acid sequence into a protein sequence; And generating an initial population set based on the protein sequence.

In some embodiments, the initial population set is of a predetermined size.

In some embodiments, the initial population set comprises a binary representation of a plurality of initial candidate nucleic acid sequences.

In some embodiments, performing optimization of the harmonic index, codon context index, and outlier index comprises: maximizing the harmonic index; Maximizing the codon context index; And minimizing the outlier index.

In some embodiments, the step of performing optimization of the harmonic index, codon context index, and outlier index comprises, for each initial candidate nucleic acid sequence of the initial population set, each harmonic index value for each initial candidate nucleic acid sequence, each Calculating a codon context index value and each outlier index value; Allocating a plurality of suitability values corresponding to a plurality of initial candidate nucleic acid sequences based on the calculating step; Classifying a plurality of initial candidate nucleic acid sequences based on the plurality of suitability values; And including a subset of the sorted plurality of initial candidate nucleic acid sequences in a subsequent population set. In some embodiments, the plurality of suitability values comprises a harmony index, a codon context index, and an outlier index for a candidate nucleic acid sequence.

In some embodiments, the method further comprises generating a progeny population based on the initial population; And including the progeny population in the subsequent population set.

In some embodiments, the progeny population is generated through binary tournament selection, cross/recombination, mutation, or any combination thereof.

In some embodiments, the initial population set and the subsequent population set are the same size.

In some embodiments, the step of performing optimization of the harmonic index, codon context index, and outlier index comprises a plurality of iterations, wherein the i-th iteration of the plurality of iterations is: a nucleic acid corresponding to the (i-1)th iteration Receiving a population set of sequences; associating each nucleic acid sequence of the population set corresponding to the (i-1) th iteration with a non-dominant level; Classifying nucleic acid sequences in the population set corresponding to the (i-1) th iteration based on the associated non-dominant level; generating a population set corresponding to the i-th iteration, wherein the population set corresponding to the i-th iteration is a subset of classified nucleic acid sequences corresponding to the (i-1)th iteration and the (i-1)th iteration Comprising a progeny population generated based on the classified nucleic acid sequence corresponding to the repeat; And determining whether to proceed to the (i+1) th iteration using the population set corresponding to the i-th iteration, based on the one or more termination conditions.

In some embodiments, associating each nucleic acid sequence with a non-dominant level comprises, for each nucleic acid sequence in the population set corresponding to the (i-1) th iteration, each harmony index value, each codon context index value, and And calculating each outlier index value.

In some embodiments, generating a population set corresponding to the i-th iteration comprises associating at least one nucleic acid sequence of the classified nucleic acid sequence corresponding to the (i-1)th iteration with one of a plurality of predetermined reference points. Includes steps.

In some embodiments, one or more termination conditions reach a fixed number of iterations and reach the best fit and do not produce better results, the minimum criterion of a solution that is close to optimal is satisfied by some solutions, or any of these. Includes a combination of.

In some embodiments, the harmony index of the candidate nucleic acid sequence is calculated based on the formula: H = 1-D( F _hs ,F _ts ), wherein D() represents a distance function; Wherein F _hs comprises a vector comprising the frequency of synonym codons of a plurality of amino acids in a plurality of highly expressed genes; F _ts includes a vector containing the frequency of synonym codons of a plurality of amino acids in the coding gene of the candidate nucleic acid sequence.

In some embodiments, D() represents a function that measures the distance between two vectors. In some embodiments, D() is a distance function including, but not limited to, a Euclidean distance, a cosine distance, a Manhattan distance, or a Minkowski distance of two vectors.

In some embodiments, the frequency of synonym codons of a plurality of highly expressed genes or candidate nucleic acid sequences is defined as follows:

In some embodiments, the codon context index of the candidate nucleic acid sequence is calculated based on the formula: CC = 1 -D( F _hcc ,F _tcc ), wherein D() represents a distance function; Wherein F _hcc comprises a vector comprising the frequency of synonym codon pairs of two consecutive amino acids in a plurality of highly expressed genes; F _tcc comprises a vector comprising the frequency of synonym codon pairs of two consecutive amino acids within the coding gene of the candidate nucleic acid sequence.

In some embodiments, D() represents a function that measures the distance between two vectors. In some embodiments, D() is a distance function including, but not limited to, a Euclidean distance, a cosine distance, a Manhattan distance, or a Minkowski distance of two vectors.

In some embodiments, the frequency of synonym codon pairs of a plurality of highly expressed genes or candidate nucleic acid sequences is defined as follows:

에 기초하여 계산되며, 상기에서 N은 복수의 미리결정된 서열 특징의 수이고; 여기서 fi(x)는 복수의 미리결정된 서열 특징 중 i번째 서열 특징의 페널티 스코어링 함수를 나타내고; wi는 fi(x)와 연관된 상대적 가중치를 나타낸다.In some embodiments, the outlier exponent is

Where N is the number of a plurality of predetermined sequence features; Where fi(x) represents a penalty scoring function of the i-th sequence feature among a plurality of predetermined sequence features; wi denotes a relative weight associated with fi(x).

In some embodiments, the plurality of predetermined characteristics comprises a GC-content value, a CIS element, a repeat element, an RNA splicing site, a ribosome binding sequence, a minimum free energy of an mRNA, or any combination thereof.

In some embodiments, a plurality of predetermined features are identified based on the selected expression system.

In some embodiments, the modification of the NSGA-III algorithm comprises an EliteNSGA-III algorithm or an NSGA-II based immune algorithm.

In some embodiments, the step of performing optimization of the harmonic index, codon context index, and outlier index comprises a plurality of optimized values by descending order of harmonic index, then descending order of codon context index, and then ascending order of outlier index. Ranking the nucleic acid sequences; Selecting one or more topmost optimized nucleic acid sequences for synthesis.

In some embodiments, the method further comprises c) removing a predetermined adverse subsequence or motif from the optimized nucleic acid sequence of the plurality of optimized nucleic acid sequences.

In some embodiments, a predetermined adverse subsequence or motif is identified based on analysis of a plurality of text portions.

In some embodiments, removing the predetermined adverse subsequence or motif comprises identifying a predetermined adverse subsequence or motif in the optimized nucleic acid sequence; Identifying a plurality of synonym codons based on the identified predetermined adverse subsequence or motif; Selecting a synonym codon from the plurality of synonym codons for substitution with a predetermined adverse subsequence identified in the optimized nucleic acid sequence.

In some embodiments, at least one of the harmonic index, codon context index, and outlier index is calculated based on one or more characteristics of the plurality of highly-expressed genes from one or more databases.

In some embodiments, the one or more characteristics include codon frequencies, synonym codon frequencies, codon pair frequencies, or combinations thereof.

In some embodiments, the method further comprises setting one or more parameters, wherein the one or more parameters are the size of the population set, the number of divisions, the distribution index for the simulated binary intersection, the simulated binary intersection. The crossover ratio for the bit flip mutation, the mutation ratio for the bit flip mutation, the distribution index for the bit flip mutation, or any combination thereof.

In some embodiments, a non-transitory computer-readable storage medium is provided that stores one or more programs, the one or more programs being executed by one or more processors of the electronic device, the electronic device using any of the methods described herein. Contains commands that cause them to perform.

In some embodiments, a system is provided for optimizing a nucleic acid sequence for protein expression in a host, the system comprising: one or more processors; Memory; And one or more programs, wherein the one or more programs are stored in memory and configured to be executed by one or more processors, the one or more programs comprising instructions for performing any of the methods described herein.

In some embodiments, an electronic device is provided for optimizing a nucleic acid sequence for expression of a protein in a host, the device comprising means for performing any of the methods described herein.

In some embodiments, a program product stored in a recordable medium is provided for optimizing a nucleic acid sequence for expression of a protein in a host, the program product comprising computer software for performing any of the methods described herein.

In some embodiments, an isolated nucleic acid molecule is provided comprising an optimized nucleic acid sequence obtained from any of the methods described herein.

In some embodiments, a vector comprising the above-mentioned isolated nucleic acid molecule is provided.

In some embodiments, a recombinant host cell comprising the above-mentioned isolated nucleic acid molecule or above-mentioned vector is provided.

In some embodiments, a method for expressing a protein in a host cell is provided, the method comprising: (a) obtaining an optimized nucleic acid sequence for expression of the protein in a host cell using any of the methods described herein. Step, (b) synthesizing a nucleic acid molecule comprising the optimized nucleic acid sequence; (c) introducing the nucleic acid molecule into a host cell to obtain a recombinant host cell; And (d) culturing the recombinant host cell under conditions that allow expression of the protein from the optimized nucleic acid sequence.

1 shows a block diagram of an exemplary process for codon optimization, in accordance with some embodiments.
2A depicts an exemplary pipeline for building and executing an algorithm for optimizing a sequence (eg, a nucleic acid sequence) for expression of a protein in a host, in accordance with some embodiments.
2B shows an exemplary general workflow of a genetic algorithm, in accordance with some embodiments.
3 shows Western blot results of GFP and JNK3A1 optimized compared to their wild type, according to some embodiments.
4 illustrates an exemplary electronic device, in accordance with some embodiments.

The present invention provides improved codon optimization to improve the recombinant expression of genes in a variety of hosts including, but not limited to, E. coli, CHO, HEK293, yeast, insect, cell-free expression systems, and the like. Exemplary systems according to the invention collect highly-expressed genes for expression systems, extract basic sequence features, replicate beneficial generic patterns in sequences of interest (e.g., nucleic acid sequences), and in expression systems Adverse features are removed to improve the expression of the target gene.

Currently, a number of codon optimization tools have been developed and are summarized in Table 1 below. Codon usage (eg, codon adaptation index [CAI], number of valid codons [ENc], relative synonym codon usage [RSCU] and synonym codon usage sequence [SCUO]), codon pairs, tRNA usage (e.g., tRNA adaptation Index [tAI]), GC-content, ribosome binding site (RBS), hidden stop codons, motif avoidance, restriction site removal, mRNA secondary structure of the gene (eg, mRNA free energy) and hydrotherapy index optimization. These tools have been considered to elevate expression during codon optimization of bacteria, yeast, insect and mammalian cells, of many, preferably most or all, of these parameters and factors.

Table 1

However, since too many factors can be taken into account for the point, how to balance them is a multiple target optimization problem, but it is difficult because the targets can conflict with each other. On the other hand, omission of one or more factors or parameters from consideration can result in low or no expression of the target gene in the expression system.

Systems and methods are provided herein for improved codon optimization that takes into account as well as balances multiple factors using a multi-goal optimization algorithm. According to some embodiments, codon optimization has, among other things, three goals: (i) how to initially allocate the coefficients of synonym codons for a particular amino acid, (ii) how to place synonym codons at the most suitable positions of the synonym codons, And (iii) a method of reducing unfavorable but accidentally generated subsequences and/or motifs. In some embodiments, these three goals are quantified as a harmonic index, a codon context index, and an outlier index. During optimization, targets are considered using a multi-goal algorithm such as Non-Dominant Classification Genetic Algorithm III (NSGA-III) or a variant thereof. Specifically, a target can be calculated for a given candidate nucleic acid sequence with reference to the known properties of the highly-expressed gene. In some embodiments, various known adverse motifs and/or features (eg, as identified in the literature) are removed from one or more optimized sequences prior to gene synthesis and protein expression.

Accordingly, the present invention is preferably thereby codon matching, codon usage (e.g., synonymous codon distribution), codon context index, cis-acting mRNA destabilization motif, RNase splicing site, GC-content, ribosome binding site. (RBS), the mRNA secondary structure of the gene (e.g., mRNA free energy) and all or most of the parameters and factors affecting protein expression including, but not limited to, repeating elements are mammals, insects, yeast, Considerations for improving and optimizing nucleic acid sequences to elevate protein expression of genes in expression host cells, including both eukaryotic and prokaryotic cells, such as bacteria and algae, and in expression systems, such as in cell-free expression systems. It provides a systematic way to be taken on.

Thus, in one aspect, the invention provides a method for sequence optimization for improved recombinant protein expression using the NSGA-III algorithm or modifications thereof to optimize multiple (e.g., more than two) targets. . In another aspect, methods are provided for removing adverse motifs and features from nucleic acid sequences prior to gene synthesis and protein expression (eg, after repetition of the NSGA-III algorithm has been completed). Also provided are methods for quantifying and calculating multiple targets in an optimization algorithm, as well as identifying adverse motifs and features to reduce or eliminate.

Also provided are systems, non-transitory computer-readable storage media, electronic devices, and program products for storing one or more programs for performing any one or more steps of any of the methods described herein. An isolated nucleic acid molecule comprising an optimized nucleic acid sequence obtained from the methods described herein; A vector containing the isolated nucleic acid molecule; Also provided is a recombinant host cell comprising the isolated nucleic acid molecule or the vector. Also provided are methods of expressing a protein in a host cell, including any of the methods described herein.

It is understood that the embodiments of the invention described herein include “consisting of” and/or “consisting essentially of” of the embodiments.

Reference to "about" for a value or parameter in this specification includes (and describes) modifications to the value or parameter itself. For example, description referring to “about X” includes description of “X”.

As used herein, reference to “not” to a value or parameter generally means and describes “other than” a value or parameter. For example, that the method is not used to treat type X cancer means that the method is used to treat type X cancer.

As used in this specification and the appended claims, the singular forms “a”, “or” and “the” include plural referents unless the context clearly dictates otherwise.

As used in this specification and the appended claims, “set” refers to one or more objects unless the context clearly dictates otherwise.

Method of codon optimization

In one aspect, the invention provides a method of optimizing a nucleic acid sequence for expression of a protein in a host (eg, computer-implemented or computer-assisted method). Methods of removing adverse motifs and features from nucleic acid sequences prior to gene synthesis and protein expression (eg, after repetition of the NSGA-III algorithm has been completed) are relevant for these methods. Also relevant to these methods is how to quantify and calculate multiple targets in the optimization algorithm, as well as how to identify adverse motifs and features to reduce or eliminate.

1 illustrates an exemplary process 100 for codon optimization with dash blocks representing optional steps. While some of the process 100 is described herein as being performed by a particular device, it will be understood that the process 100 is not so limited. In other examples, process 100 is performed using only a single electronic device (eg, electronic device 400) or multiple electronic devices. In process 100, some blocks are selectively combined, the order of some blocks is selectively changed, and some blocks are selectively omitted. In some examples, additional steps may be performed in combination with process 100.

At block 106, the electronic device receives an initial population set, wherein the initial population set includes a plurality of initial candidate nucleic acid sequences capable of expressing a protein. In some embodiments, the initial population set is randomly generated. In some embodiments, the initial population set is of a predetermined size (eg, determined by the user).

In some embodiments, as shown at block 106, receiving the initial population set includes generating an initial population set based on the protein sequence. For example, receiving an initial population set may include receiving a protein sequence (eg, as input from a user); And generating an initial population set based on the received protein sequence. As another example, receiving an initial population set may include receiving a nucleic acid sequence (eg, as input from a user); Translating the received nucleic acid sequence into a protein sequence; And generating an initial population set based on the protein sequence.

In some embodiments, the initial population set comprises a binary representation of a plurality of initial candidate nucleic acid sequences (eg, binary strings). In general, a binary string that is not a codon list/array/vector is selected as a data structure representing a coding gene, and all manipulation targets of a genetic algorithm including population initialization, cross/recombination, mutation, and selection evaluate the suitability of the gene before selection. It is a binary string except for. As described further below, in some embodiments, when a fit function (i.e., three exponential functions) needs to be evaluated for each individual of the entire population prior to selection, the binary representation is temporarily reverted to the codon string. It should be converted.

In block 108, the electronic device performs optimization of the harmonic index, codon context index, and outlier index using the computer-assisted NSGA-III algorithm or a variant thereof based on the initial population set, whereby the protein may be expressed. A number of optimized nucleic acid sequences are obtained.

Always or in some embodiments, the harmony index of the candidate nucleic acid sequence indicates the consistency of the distribution of the frequency of use of the synonym codon between the plurality of highly expressed genes and the candidate nucleic acid sequence (i.e., the gene encoding the candidate protein during optimization), which It helps to solve how to assign the coefficients of synonym codons for specific amino acids. The codon context index of a candidate nucleic acid sequence is a measure for placing synonym codons in appropriate positions. The outlier index of a candidate nucleic acid sequence is a measure of the negative effect of a plurality of predetermined sequence features on the candidate nucleic acid sequence.

In some embodiments, as shown in block 106, performing optimization of a harmonic index, a codon context index, and an outlier index comprises: maximizing the harmonic index; Maximizing the codon context index; And minimizing the outlier index.

Optimization can be performed by using a multi-goal genetic algorithm, the three goals are maximizing the harmonic index; Maximizing the codon context index; And to minimize the outlier index. In some embodiments, the NSGA-III algorithm or variant is used. Unlike traditional genetic algorithms, the maintenance of diversity among population members of NSGA-III is supported by providing and adaptively updating a number of well-distributed predetermined reference points, so NSGA-III has significant changes in its selection manipulators. There is. Additionally, NSGA-III demonstrates its efficacy in solving the three-goal to fifteen-goal optimization problem compared to other genetic algorithms such as NSGA-II. Variations of the NSGA-III algorithm include the EliteNSGA-III algorithm, the NSGA-II based immune algorithm, MAM-MOIA or MOLA. The EliteNSGA-III algorithm is described in a publication entitled "ELITENSGA-III: AN IMPROVED EVOLUTIONARY MANY-OBJECTIVE OPTIMIZATION ALGORITHM" published in 2016 by Amin Ibrahim et al., which is incorporated herein by reference in its entirety. Various immune algorithms are, for example, a publication entitled "MOIA: MULTI-OBJECTIVE IMMUNE ALGORITHM" published in September 2010 by Guan-Chun Luh et al. -OBJECTIVE OPTIMIZATION", published in April 2010 by Zhi-Hua Hu under the heading "A MULTIOBJECTIVE IMMUNE ALGORITHM BASED ON A MULTIPLE-AFFINITY MODEL" and Chinese Patent Application No. 201710611752.5 filed July 25, 2017 And are incorporated herein by reference in their entirety.

According to the operation of the NSGA-III algorithm (or similar genetic algorithm), the step of performing optimization of the harmonic index, codon context index, and outlier index is for each initial candidate nucleic acid sequence of the initial population set, each initial candidate nucleic acid. Calculating each harmonic index value, each codon context index value, and each outlier index value for the sequence; Allocating a plurality of suitability values corresponding to a plurality of initial candidate nucleic acid sequences based on the calculating step; Classifying a plurality of initial candidate nucleic acid sequences based on the plurality of suitability values; And incorporating a subset of the aligned plurality of initial candidate nucleic acid sequences into a subsequent population set (ie, used in a second iteration).

In accordance with the operation of the NSGA-III algorithm (or similar genetic algorithm), the method comprises the steps of generating a progeny population based on the initial population; And including the progeny population in a subsequent population set (ie, used in the second iteration). In some embodiments, the progeny population is generated through binary tournament selection, cross/recombination, mutation, or any combination thereof.

In some embodiments, the initial population set and the subsequent population set (ie, used in the second iteration) are of the same size.

Depending on the operation of the NSGA-III algorithm (or similar genetic algorithm), the step of performing optimization of the harmonic index, codon context index and outlier index includes a plurality of iterations. The i-th iteration of the plurality of iterations (where i can be 2, 3, 4, 5, 6...n) comprises: receiving a population set of nucleic acid sequences corresponding to the (i-1)th iteration ; associating each nucleic acid sequence of the population set corresponding to the (i-1) th iteration with a non-dominant level; Classifying nucleic acid sequences in the population set corresponding to the (i-1) th iteration based on the associated non-dominant level; generating a population set corresponding to the i-th iteration, wherein the population set corresponding to the i-th iteration is a subset of the classified nucleic acid sequence corresponding to the (i-1)th iteration and the (i-1)th iteration Comprising a progeny population generated based on the classified nucleic acid sequence corresponding to the repeat; And determining whether to proceed to the (i+1) th iteration using the population set corresponding to the i-th iteration, based on the one or more termination conditions.

In some embodiments, associating each nucleic acid sequence with a non-dominant level comprises, for each nucleic acid sequence in the population set corresponding to the (i-1) th iteration, each harmony index value, each codon context index value, and Each outlier index value is included.

In accordance with the operation of the NSGA-III algorithm, in some embodiments, generating a population set corresponding to the i-th iteration comprises a plurality of at least one nucleic acid sequence of the classified nucleic acid sequence corresponding to the (i-1)th iteration. And associating with one of the predetermined reference points.

In some embodiments, one or more termination conditions reach a fixed number of iterations and reach the best fit and do not produce better results, the minimum criterion of a solution that is close to optimal is satisfied by some solutions, or any of these. Includes a combination of.

In some embodiments, the method further comprises setting one or more parameters for the optimization algorithm, wherein the one or more parameters are the size of the population set, the number of partitions, the distribution index for the simulated binary intersection, the simulation The crossover ratio for the resulting binary crossover, the mutation ratio for the bit flip mutation, the distribution index for the bit flip mutation, or any combination thereof.

In some embodiments, during optimization at least one of a harmonic index, a codon context index, and an outlier index is calculated based on one or more characteristics of a plurality of highly-expressed genes from one or more databases. In some embodiments, the one or more characteristics include codon frequencies, synonym codon frequencies, codon pair frequencies, or combinations thereof. These properties of highly-expressed genes can be used to calculate the harmonic index, codon context index, and outlier index for a given candidate nucleic acid sequence as shown in the equation below.

In some embodiments, as indicated at block 102, these properties of the highly-expressed gene are identified based on private or public databases. For example, the database(s) may be a proprietary database containing previously successfully optimized orders collected in the company's order system. As another example, data may be obtained by a method of data mining of RNA-seq data under various culture conditions, which may be public information. Data processing is performed for the purpose of obtaining basic information of highly-expressed genes including codon frequencies, synonym codon frequencies, and codon pair frequencies.

In some embodiments, the harmony index of the candidate nucleic acid sequence is calculated based on the formula: H = 1-D( F _hs ,F _ts ), wherein D() represents a distance function; Wherein F _hs comprises a vector comprising the frequency of synonym codons of a plurality of amino acids in a plurality of highly expressed genes; F _ts includes a vector containing the frequency of synonym codons of a plurality of amino acids in the coding gene of the candidate nucleic acid sequence.

In some embodiments, D() represents a function that measures the distance between two vectors. In some embodiments, D() is a distance function including, but not limited to, a Euclidean distance, a cosine distance, a Manhattan distance, or a Minkowski distance of two vectors.

In some embodiments, the frequency of synonym codons of a plurality of highly expressed genes or candidate nucleic acid sequences is defined as follows:

In some embodiments, the codon context index of the candidate nucleic acid sequence is calculated based on the formula: CC = 1 -D( F _hcc ,F _tcc ), wherein D() represents a distance function; Wherein F _hcc comprises a vector comprising the frequency of synonym codon pairs of two consecutive amino acids in a plurality of highly expressed genes; F _tcc comprises a vector comprising the frequency of synonym codon pairs of two consecutive amino acids within the coding gene of the candidate nucleic acid sequence.

In some embodiments, D() represents a function that measures the distance between two vectors. In some embodiments, D() is a distance function including, but not limited to, a Euclidean distance, a cosine distance, a Manhattan distance, or a Minkowski distance of two vectors.

In some embodiments, the frequency of synonym codon pairs of a plurality of highly expressed genes or candidate nucleic acid sequences is defined as follows:

에 기초하여 계산되며, 상기에서 N은 복수의 미리결정된 서열 특징의 수이고; 여기서 fi(x)는 복수의 미리결정된 서열 특징 중 i번째 서열 특징의 페널티 스코어링 함수를 나타내고; wi는 fi(x)와 연관된 상대적 가중치를 나타낸다.In some embodiments, the outlier exponent is

Where N is the number of a plurality of predetermined sequence features; Where fi(x) represents a penalty scoring function of the i-th sequence feature among a plurality of predetermined sequence features; wi denotes a relative weight associated with fi(x).

In some embodiments, the plurality of predetermined characteristics comprises a GC-content value, a CIS element, a repeat element, an RNA splicing site, a ribosome binding sequence, a minimum free energy of an mRNA, or any combination thereof.

In some embodiments, a plurality of predetermined features are identified based on the selected expression system. For various expression systems, the list of adverse factors may change, and their influences or weights are also not equivalent.

In some embodiments, the step of performing optimization of the harmonic index, codon context index, and outlier index comprises a plurality of optimized values by descending order of harmonic index, then descending order of codon context index, and then ascending order of outlier index. Ranking the nucleic acid sequences; Selecting one or more topmost optimized nucleic acid sequences for synthesis.

At block 110, the method further comprises c) removing a predetermined adverse subsequence or motif from the optimized nucleic acid sequence of the plurality of optimized nucleic acid sequences. In some embodiments, removing the predetermined adverse subsequence or motif comprises identifying a predetermined adverse subsequence or motif in the optimized nucleic acid sequence; Identifying a plurality of synonym codons based on the identified predetermined adverse subsequence or motif; Selecting a synonym codon from the plurality of synonym codons for substitution with a predetermined adverse subsequence identified in the optimized nucleic acid sequence.

In some embodiments, as indicated at block 104, a predetermined adverse subsequence or motif is identified based on an analysis of a plurality of text portions (eg, automatic text mining or manual inspection of the document).

In some embodiments, the method further comprises providing an output representative of one or more of the plurality of optimized nucleic acid sequences.

In some embodiments, a non-transitory computer-readable storage medium is provided that stores one or more programs, the one or more programs being executed by one or more processors of the electronic device, the electronic device using any of the methods described herein. Contains commands that cause them to perform.

In some embodiments, a system is provided for optimizing a nucleic acid sequence for expression of a protein in a host, the system comprising: one or more processors; Memory; And one or more programs, wherein the one or more programs are stored in memory and configured to be executed by one or more processors, the one or more programs comprising instructions for performing any of the methods described herein.

In some embodiments, an electronic device is provided for optimizing a nucleic acid sequence for expression of a protein in a host, the device comprising means for performing any of the methods described herein.

In some embodiments, a program product stored in a recordable medium is provided for optimizing a nucleic acid sequence for expression of a protein in a host, the program product comprising computer software for performing any of the methods described herein.

In some embodiments, an isolated nucleic acid molecule is provided comprising an optimized nucleic acid sequence obtained from any of the methods described herein.

In some embodiments, a vector comprising the above-mentioned isolated nucleic acid molecule is provided.

In some embodiments, a recombinant host cell comprising the above-mentioned isolated nucleic acid molecule or above-mentioned vector is provided.

In some embodiments, a method for expressing a protein in a host cell is provided, the method comprising the steps of: (a) obtaining an optimized nucleic acid sequence for expression of the protein in the host cell using any of the methods described herein. , (b) synthesizing a nucleic acid molecule comprising the optimized nucleic acid sequence; (c) introducing the nucleic acid molecule into a host cell to obtain a recombinant host cell; And (d) culturing the recombinant host cell under conditions that allow expression of the protein from the optimized nucleic acid sequence.

2A illustrates an exemplary pipeline 200 for building and executing an algorithm for optimizing a sequence (eg, a nucleic acid sequence) for expression of a protein in a host in accordance with some embodiments of the present invention. Process 200 is performed, for example, using one or more electronic devices shown in FIG. 4. In some examples, process 200 is performed using a client-server system, and the block of process 200 is partitioned in any way between the server and the client device. In another example, the block of process 200 is partitioned between a server and/or multiple client devices. Accordingly, while portions of process 200 are described herein as being performed by a particular device, it will be appreciated that process 200 is not so limited. In other examples, process 200 is performed using only a single electronic device (eg, electronic device 400) or multiple electronic devices. In process 200, some blocks are selectively combined, the order of some blocks is selectively changed, and some blocks are optionally omitted. In some examples, additional steps may be performed in combination with process 200.

Data collection and literature review

2A, at block 202, a plurality of highly-expressed genes may be identified from one or more databases. The database can be public or private. For example, the database(s) may be a proprietary database containing previously successfully optimized orders collected in the company's order system. As another example, data may be obtained by a method of data mining of RNA-seq data under various culture conditions, which may be public information.

In block 204, the basic characteristics of the highly-expressed gene are identified. In an exemplary embodiment, mRNA-seq experiments and data analysis are performed according to Illumina's recommended mRNA-Seq workflow for standard samples. During the process, the TruSeq Stranded mRNA Library Prep Kit can be used for library preparation, and NextSeq's PE300 can be used for sequencing. In succession, data processing through TopHat, Cufflinks and homemade scripts can be applied to obtain basic information of highly-expressed genes including codon frequencies, synonym codon frequencies, and codon pair frequencies.

At blocks 206 and 208, the example system may also identify any reported and verified adverse features to avoid in order to maintain the established advantage. In order to find negative factors that can lead to a decrease in protein expression, the system can perform a literature review. For example, expression-related adverse motifs and mRNA characteristics reported by the method of automatic text mining and/or manual inspection can be identified for a variety of hosts.

Key factors/suitability functions for optimization algorithms

Expression of the coding gene has a number of steps depending on the level of transcription, mRNA conversion, translation (including initiation, promoter escape, extension and termination) and post-translational modification. Nevertheless, codon optimization is simplified to a problem of combination and has three intuitive manipulations: (i) how to initially assign the synonym codon coefficient of a specific amino acid, and (ii) how to place the synonym codon at the most suitable position of the synonym codon. , And (iii) unfavorable but accidentally generated subsequences and/or motifs.

According to some embodiments of the present invention, the three key factors that correspond to each of the three above-mentioned manipulations and have a high correlation with protein expression are provided below: the harmonic index, the codon context index, and the outlier index. As discussed below, these three indicators are calculated based on the above-mentioned basic data collected from various data sources.

Referring to FIG. 2A, at block 210, an optimization procedure comprising two steps 212 and 214 is performed. In step 1 shown in block 212, the system performs a multi-target codon optimization based on the NSGA-III algorithm or a variant thereof, which maximizes the harmonic index, the codon context index, and the outlier index. It includes the step of minimizing.

1. Harmony index

The harmony index represents the consistency of the distribution of the frequency of use of synonym codons between the highly expressed gene and the candidate nucleic acid sequence. A candidate nucleic acid sequence refers to a gene encoding a candidate protein evaluated in at least one iteration of the optimization algorithm, which is described in detail under the heading “Multi-Object Optimization Algorithm”. In some embodiments, the harmonic index is defined as follows:

In the above equation, H is a harmonic index, and D() is a distance function between two vectors, which may be a Euclidean distance, a cosine distance, a Manhattan distance, or a Minkowski distance, but is not limited thereto. F _hs is a vector containing the frequency of synonym codons of 18 amino acids (except Met/M and Trp/W) in a highly expressed gene, 3 stop codons (i.e., TAA, TAG and TGA), amino acids from 64 codons It contains 59 elements due to the removal of the codon of Met/M (i.e. ATG) and the codon of amino acid Trp/W (i.e. TGG). F _ts is a vector containing the frequency of the synonym codon of 18 amino acids in the coding gene (ie, candidate nucleic acid sequence) of the candidate protein awaiting codon optimization.

Compared to the Codon Adaptation Index (CAI), the Harmonic Index focuses on the distribution of synonym codons (i.e. used balance/charging balance), but always the maximum CAI through the step of choosing the best one most frequently occurring synonym codon only It is not aimed at.

In some embodiments, the frequency of a particular synonym codon of a highly expressed gene or candidate nucleic acid sequence used during calculation of the harmony index is defined as follows:

Harmony indices take into account codon use, but are only concerned with the frequency distribution of synonym codons, while their assignment to different loci of one of the 18 amino acids is still a problem (i.e., ordering of synonym codons of the same amino acid). Therefore, the codon context index described below is required to resolve this bottleneck through synonym codon matching to select the roughly optimal ranking for the synonym codon.

2. Codon context index

The codon context index of a candidate nucleic acid sequence is a measure for placing synonym codons in appropriate positions. In some embodiments, the codon context index is defined as follows:

In the above equation, CC denotes a codon context index, and D() is a distance function between two vectors, which may be a Euclidean distance, a cosine distance, a Manhattan distance, or a Minkowski distance, but is not limited thereto. F _hcc is a vector containing the frequencies of synonym codon pairs of two consecutive amino acids of all kinds within a highly expressed gene. For example, the amino acid Phe/F has two synonymous codons, TTT and TTC; The amino acids Lys/K also have AAA and AAG as codons; Their synonym codon pair must be a 2 x 2 combination, including TTTAAA, TTTAAG, TTCAAA and TTCAAG. Since there are no synonymous codon pairs for the permutations of the two amino acids methionine/M and tryptophan/W (i.e. MM, MW, WW and WM), the length of CC is 61 x 61 minus 4 and finally 3717. F _tcc is a vector containing the frequency of synonym codon pairs of two consecutive amino acids of all kinds within the coding gene of the candidate protein (i.e., the candidate nucleic acid sequence), the length of which is also 3717.

The frequency of a particular synonym codon pair of a highly expressed gene or candidate nucleic acid sequence used during the calculation of the codon context index is defined as follows:

3. Outlier Index

Outlier index is a measure calculated by a weighting function to assess the negative impact of a plurality of identified sequence features on protein expression. In some embodiments, the outlier index is defined as follows:

In the above formula, N is the number of a plurality of identified sequence factors and N>1. f _i (x) represents the penalty scoring function of the i-th sequence factor of the identified N sequence feature; And wi denotes the relative weight given to _{f i (x).} Therefore, the optimized gene should have the lowest outlier index.

In some embodiments, a plurality of sequence factors may be identified through one or more of steps 202, 204, and 208 shown in FIG. 2A. In some embodiments, the plurality of sequence factors contain, but are not limited to, the GC-content, CIS elements, repetitive elements, RNA splicing sites, ribosome binding sequences, minimal free energy of mRNA, detailed below.

3(a). mRNA Minimum free energy ( MFE )

Potentially strong stem-loop secondary structure of mRNA located downstream of the initiation codon can interfere with the migration of the ribosome complex, thus slowing translation and reducing translation efficiency. The stable secondary structure of the mRNA may cause the ribosome complex to dislodge from the mRNA and lead to premature termination of translation. There are several methods for free energy calculation and secondary structure prediction, including mfold, RNAfold and RNAstructure. According to an embodiment of the invention, the local secondary structure of an mRNA with a low free energy (ΔG <-18 Kcal/mol) or a long complementary stem (> 10 bp) is defined as too stable for efficient translation. The gene sequence is preferably optimized so that the local structure is not so stable. Both of the 5'-UTR and 3'-UTR of the mRNA are preferably taken into account for the calculation of the mRNA structure free energy and prediction of the secondary structure.

In some embodiments, secondary structures that are considered too stable are associated with a higher penalty. The weights used to give a higher penalty score are flexible.

3(b). GC -content

The GC content of the mRNA is also preferably considered. The ideal range for GC% is approximately 30-70%. The high GC-content will allow the mRNA to form a strong stem-loop secondary structure. It will also cause problems with PCR amplification and gene cloning. The high GC-content of the target sequence is mutated (e.g., during operation of the NSGA-III algorithm, including crossover and mutation of binary strings) using codon weights to be preferably about 50-60%.

There are two different measures of GC%. One is the global GC% averaged along the entire sequence; The other is more useful, which is the local GC% calculated within a fixed size (eg, 60 bp) shifted “window”. According to an embodiment of the present invention, the local GC% is optimized to about 35-65%.

3(c). Unstable factor (for example, Cis -Action mRNA Destabilization motif, RNase Splicing sites and repeating elements, etc.)

Cis-acting mRNA destabilization motifs including, but not limited to, AU-rich elements (ARE) and RNase recognition and cleavage sites, in order to reduce or minimize mRNA degradation or increase the stability of the mRNA and thus reduce the conversion time of the mRNA It is preferably mutated or deleted from the gene sequence. The AU-rich element (ARE) with the core motif of AUUUA (SEQ ID NO: 1) is generally found in the 3'untranslated region of the mRNA. Another example of an mRNA cis-element consists of the sequence motif TGYYGATGYYYYY (SEQ ID NO: 2), where Y represents T or C. RNase recognition sequences include, but are not limited to, RNase E recognition sequences. Host strains with insufficient RNase can also be used for protein expression.

The RNase splicing site allows RNA splicing to produce different mRNAs and thus can reduce the original mRNA level. The RNase splicing site is also preferably non-functionally mutated to maintain the mRNA level.

In order to produce high levels of mRNA, an optimal transcriptional promoter sequence is preferably used for the gene sequence. For prokaryotic hosts such as E. coli, one of the strong promoters is the T7 promoter for T7 RNA polymerase (T7 RNAP). Some bases of long or short tandem simple sequence repeats (SSRs) are preferably mutated using codon degeneration to break the repeats to reduce polymerase slippage, thereby reducing premature protein or protein mutations.

There are additional factors and parameters that affect the level of mRNA translation and the resulting protein expression. These factors influence the translation from the initiation of the translation to the end of the translation. Ribosomes initiate translation by binding mRNA at the ribosome binding site (RBS). Since ribosomes do not bind double-stranded RNA, the local mRNA structure around this region is preferably single-stranded and does not form any stable secondary structure. AGGAGG (SEQ ID NO: 3), a consensus RBS sequence for prokaryotic cells such as E. coli, also called the Shine-Dalgarnon sequence, is preferably located at several bases just before the translation initiation site of the gene to be expressed. However, the internal ribosome entry site (IRES) is preferably mutated to prevent ribosome binding to avoid non-specific translation initiation.

Detailed descriptions of the above-mentioned factors can be found, for example, in a publication entitled "CIS/TRANSGENE OPTIMIZATION: SYSTEMATIC DISCOVERY OF NOVEL GENE EXPRESSION USING BIOINFORMATICS AND COMPUTATIONAL BIOLOGY APPROACHES" published in May 2018, 2014 A publication titled "AU-RICH ELEMENTS AND THE CONTROL OF GENE EXPRESSION THROUGH MRNA STABILITY" by Timothy J Gingerich, published in July, and "ARED-PLUS: AN UPDATED AND EXPANDED DATABASE OF" by Tala Bakheet, published in October 2017. A publication entitled "AU-RICH ELEMENT-CONTAINING MRNAS AND PRE-MRNAS", published in 2002 by Shuang Zhang et al. entitled "IDENTIFICATION AND CHARACTERIZATION OF A SEQUENCE MOTIF INVOLVED IN NONSENSE-MEDIATED MRNA DECAY", published in 2002. A publication entitled "CORRELATIONS BETWEEN SHINE-DALGARNO SEQUENCES AND GENE FEATURES SUCH AS PREDICTED EXPRESSION LEVELS AND OPERON STRUCTURES" by Jiong Ma et al. LIBRARY FOR TUNING EXPRESSION LEVEL OF MULTIPLE GENES IN MAMMALIAN CELLS", which is incorporated herein by reference in its entirety.

For various expression systems, the list of adverse factors may change, and their influences or weights are not the same. Thus, f _i (x) and its weight can be dynamically modified for various expression systems. For example, after setting the allowable range of GC-content and MFE, the degree of'out of range' will result in penalties as a percentage. Likewise, the number of occurrences of an unstable factor can be recorded directly as a penalty score.

It should be appreciated that even if the outlier index for the candidate nucleic acid sequence is high, the candidate sequence may still have the potential for the repeats to survive to maintain the diversity of the entire population. In other words, filtering of adverse motifs/features through the outlier index is not necessary as a higher outlier index (i.e., penalty) can only result in a lower survival rate. In contrast, after the iteration of the NSGA-III algorithm is complete (i.e., in step 110 of FIG. 1 or step 214 of FIG. 2) removal of the adverse motif/feature is essential.

In conclusion, the present invention not only attempts to enhance the positive effect by maximizing the values of the harmonic index and the codon context index, but also does its best to avoid adverse effects by minimizing the outlier index.

Multi-goal (e.g., more than 2 goals) optimization algorithm

Since the present invention is an optimization task of three generic targets, a multi-target genetic algorithm can be used. In some embodiments, the NSGA-III algorithm or a variant thereof such as EliteNSGA-III (also presented by K. Deb) solves the multi-target optimization problem by maintaining population diversity during selection manipulation of the classical framework of the genetic algorithm. It can be used due to their advantage in doing so.

NSGA-III was proposed in 2014 by Kalyanmoy Deb and Himanshu Jain. It is a reference-point-based multi-goal evolutionary algorithm that is non-dominant but follows the NSGA-II framework that emphasizes population members close to a given set of reference points. NSGA-III demonstrates its efficacy in solving the 3-goal to 15-goal optimization problem compared to other genetic algorithms such as NSGA-II. Unlike conventional genetic algorithms, the maintenance of diversity among population members in NSGA-III is supported by providing and adaptively updating a number of well-diffused pre-defined reference points, and thus NSGA-III is dependent on its selective effector. There are significant changes.

The NSGA-III algorithm is described in a publication entitled "An Evolutionary Many-Objective Optimization Algorithm Using Reference-Point-Based Nondominated Sorting Approach, Part I: Solving Problems With Box Constraints" by Kalyanmoy Deb et al. published in August 2014. , Which is incorporated herein by reference in its entirety. The relevant NSGA-II algorithm is described in a publication entitled "A FAST AND ELITIST MULTIOBJECTIVE GENETIC ALGORITHM: NSGA-II" by Kalyanmoy Deb et al. published in August 2002, which is incorporated herein by reference in its entirety.

During the implementation of NSGA-III, a binary string, not a codon list/arrangement/vector, is selected as a data structure representing a nucleic acid sequence, and all general manipulation objects of the general genetic algorithm, including population initialization, cross/recombination, and mutation, are: Binary strings are binary strings because as a data structure they require less computer memory and enable faster manipulation speed compared to codon lists/arrays/vectors. In some embodiments, 3 consecutive bits are used to represent codons at one position because the number of all combinations of 3 bits is sufficient to match all possible candidates for synonym codons of a particular amino acid. For example, 3 bits have 8 kinds of combinations, e.g., 000, 001, 010, 011, 100, 101, 110, 111, and their coefficients are arbitrary amino acids, even amino acids with 6 synonym codons each Greater than the number of synonym codons for L, R and S.

Thus, each one of a 3-bit string represents a synonym codon for a given amino acid. During suitability calculations (e.g., calculation of harmonic index, codon context index, and outlier index), binary strings representing individual candidates of the population are converted back to coding sequencing (i.e., DNA). On the other hand, as discussed above, the targets of genetic algorithm manipulation (including crossover, mutation, and selection) are all binary strings, and thus the conversion is temporary. Hence, conformance calculations are based on sequence, while all other manipulations are based on binary strings for efficiency and speed.

Before the start of NSGA-III, include the size of the population, the number of divisions, the distribution index for the simulated binary intersection, the intersection ratio for the simulated binary intersection, the mutation ratio for the bit flip mutation, and the distribution index for the bit flip mutation. Thus, multiple parameters need to be set. The authors of NSGA-III propose a two-tiered approach to segmentation for multi-goal problems with a specified number of external and internal partitions. To use the two-tier approach, we could substitute the number of partitions with the number of outer partitions and the number of inner partitions. The initialization process of all individuals is random and the crossover and mutation manipulations are not significantly different from the classical genetic algorithm shown in FIG. 2B.

2B depicts an exemplary general workflow of a genetic algorithm involving bio-inspired manipulators such as selection of crossovers, mutations and population evolution. During the implementation of the present invention, a binary string represents a sequence and thus the object of all of the above operators is a binary string.

When the fit function (i.e., the three exponential functions shown previously) must be evaluated for each individual of the entire population prior to selection, the binary string is temporarily transferred back to the codon string. After multiple evolutionary generations and the end of evolution, the finally generated codon string will be chained and output to the optimal gene used for recombinant expression.

In some embodiments, the termination condition reaches a fixed number of iterations and the best fit is reached and no better results are produced, the minimum criterion for a solution that is close to optimal is satisfied by some solutions, or any combination thereof. Including, but not limited to.

According to the teachings of the NSGA-III algorithm, these optimal genes should be solutions located on the Pareto surface in three-dimensional space and treated identically. For practical purposes, due to the limited sources used in gene synthesis and expression testing, we sequence them first by descending order of harmonic index, then descending order of codon context index, and finally ascending order of outlier index. do. The best one can be selected for synthesis and the heterologous expression given a quota is only one sequence. Assuming there is no strict cost control, it is recommended to test several well-spaced candidates on the Pareto surface, e.g. one candidate with the highest harmonic index, one candidate with the highest codon context index, and one candidate with the lowest outlier index. . In the present invention, the preliminary optimal gene does not have a stop codon, so two consecutive stop codons can be attached to the 3'end of the coding sequence.

molecule Cloning Specific subsequences for removal

2A, at block 214, the optimization procedure includes the steps of motif avoidance and restriction site removal. For the purpose of enhancing the convenience of molecular cloning, some unfavorable motifs and restriction sites (eg, customer dislikes) are removed from one or more optimized sequences prior to gene synthesis and protein expression. This process includes:

Step 1: Find all subsequence positions that should be avoided.

Step 2: List all synonym codons that can be used for substitution within the subsequence.

Step 3: Synonym codons that are used more frequently within highly expressed genes have a higher priority for selection under conditions where we must ensure that no new subsequences appear at the same time.

Step 4: Iteratively process all subsequences found using Steps 2-3.

In some embodiments, as indicated by blocks 206 and 208, adverse motifs and features are identified separately for various hosts by text mining and literature review.

Exemplary realization

The exemplary realization described herein illustrates the effectiveness of the present invention for codon optimization through optimization and expression of two genes (JNK3A1 and GFP) in the CHO 3E7 cell line, the basic information of which is summarized below. In order to evaluate the expression level, a flag tag antibody was applied to perform Western blot, so a flag tag was attached to the C-terminus of the two proteins, and beta-actin was used as a loading control. Each expression experiment was repeated twice.

The mRNA-seq of CHO 3E7 cultured in several media including FreeStyle CHO expression medium and CD CHO medium (Thermofish) was performed according to the classical mRNA-seq proposal recommended by Illumina. A total of 500 sequences, integrated with our successfully optimized partial sequence, were defined as highly expressed genes of the CHO 3E7 cell line. After literature review, the following subsequences were grouped by adverse motifs, and their appearance resulted in penalties (i.e., an increase in outlier index). Appropriate local (60 bp sliding-window) and global GC-content is about 35-65%, the minimum allowable MFE ΔG of the mRNA secondary structure is -18 Kcal/mol, and outliers of these parameters cause penalties. did.

1) Splice site: GGTAAG, GGTGAT

2) AT-rich elements: ATTTTA, ATTTTTA, ATTTTTTA

3) Ribosome binding site: ACCACCATGG (SEQ ID NO: 4), GCCACCATGG (SEQ ID NO: 5)

4) Antiviral motif: TGTGT, AACGTT, CGTTCG, AGCGCT, GACGTC, GACGTT

5) CpG island: CGCGCGCG

6) Polymerase slippage site: GGGGGG, CCCCCC

7) Amyloid precursor protein 3 prime stability factor: TCTCTTTACATTTTGGTCTCTATACTACA (SEQ ID NO: 6)

8) K-box: CTGTGATA

9) Brd-box: AGCTTTA

During codon optimization through NSGA-III, the population size was set to 100 and individuals were binary encoded and randomly generated, their length equal to three times the number of amino acids in the protein, the number of evolutionary generations equal to 200,000, and the number of divisions. The number depended on the number of fit functions, the distribution index for the simulated binary intersection was 15.0, the single-point intersection ratio for the simulated binary intersection was 0.9, and the mutation ratio for the bit flip mutation was 1.0/L. , The distribution index for the bit flip mutation was 20.0.

After maximizing the harmonic index and the codon context index while minimizing the outlier index, each protein has several output optimal coding genes, of which only one gene with the greatest harmonic index was selected for the next expression test. Since EcoRI and HindIII enzymes were used for vector construction and cloning, GAATTC and AAGCTT were avoided by codon substitution.

The sequence listing submitted as an ASCII text file herein includes the optimized sequences of the two proteins GFP_Flag (SEQ ID NO: 7) and JNK3_Flag (SEQ ID NO: 8).

The detailed steps of the experiment used to evaluate the performance of the optimized gene compared to the wild type of the same gene are described below.

Step 1: Transient transfection and cell culture

1. The synthesized gene was cloned into pTT5 vector using EcoRI and HindIII enzymes. CHO 3E7 cells were cultured in FreeStyle CHO expression medium and transient transfection of the vector was performed using standard molecular biology techniques with an ^{appropriate cell-vector ratio (i.e., 1-1.2×10 6 cell density per mL for 1 ug/ml vector concentration).} Was done using.

2. After transient transfection, CHO 3E7 cells had to be cultured in suspension _{at 37°C in 5% CO 2 for 48 hours.}

Step 2 : Cell destruction

1. Take the cultured cells from the upstream and centrifuge (10,000 x g) for 2 minutes at 4°C. Discard the supernatant.

2. Resuspend the cells at the bottom of the Eppendorf tube by adding 1 mL 1*PBS. Then, centrifuge (10,000 x g) at 4° C. for 2 minutes and discard the supernatant.

3. 200 μL lysis buffer (hypopotency buffer [10mM Tris, 1.5mM MgCl ₂ , 10mM KCl, pH 7.9] + 0.5% DDM, PMSF [final concentration 1mM], nuclease, cocktail) 1*10 ⁶ Eppendorf per cell Add to the tube. Resuspend the cells with a pipette.

4. For cell destruction, put the cells in a cup-type ultrasonic cell destroyer (4℃, 3 seconds ultrasound, 1 second interval, total 10 minutes).

5. After destruction, centrifuge (12,000 x g) at 4° C. for 20 minutes. Recover the supernatant.

Step 3: sample processing

1. Measure the concentration of the supernatant using the BCA method.

2. Some of the supernatant was treated with the loading buffer.

Step 4: electrophoresis and Western Blot

1. Load the processed sample for SDS-PAGE according to the SOP. (8μg per sample)

2. After electrophoresis, Western blot experiments were performed according to SOP.

1) Transfer: Remove the gel after SDS-PAGE and transfer the protein from the gel to the PVDF membrane (transfer buffer: add 200 mL 5x transfer solution to 150 mL of absolute ethanol, dilute to 1 L, transfer for 1 hour).

2) Blocking: After transfer, PVDF was blocked with fast blocking solution for 10 minutes.

3) Incubation: After blocking, incubate for 45 minutes with 5% milk and the corresponding labeled antibody (flag tag: addition of THETM beta actin antibody, mAb, mouse GenScript, catalog number A00702 at 1:1000 dilution for 1 hour With mouse-anti-flag mAb GenScript at a dilution of 1:5000, catalog number A00187, then 1:2500 diluted labeled secondary antibody goat anti-mouse IgG-HRP GenScript, catalog number A00160 added)

4) Exposure: Exposure imaging was performed using ChemiDoc™ Touch Imaging Systems after antibody incubation and images were saved to a designated location for editing.

5) Image Lab was used for protein quantitative analysis.

3 is a Western blot result illustrating a comparison of the expression between the optimized sequence and two wild-type genes (i.e., GFP and JNK3A1) in a CHO 3E7 cell line according to an embodiment of the present disclosure, wherein only the best of each gene Only optimized solutions with high coordination index were tested for expression comparison. It has been clearly demonstrated that the present invention is effective in codon optimization and elevates expression compared to the internal control beta-actin, which is little changed. The left lane was always a ladder marker, and all expressions of a single plasmid were repeated twice. According to a rough quantitative analysis, the expression of GFP was estimated to be improved approximately 6.2-fold, and the expression of JNK3 was promoted approximately 2.4-fold after codon optimization of the present invention.

Exemplary electronic device

4 illustrates an example of a computing device according to an embodiment. The device 400 may be a host computer connected to a network. Device 400 may be a client computer or a server. As shown in Figure 4, device 400 may be any suitable type of microprocessor-based device, such as a personal computer, workstation, server, or handheld computing device (portable electronic device) such as a phone or tablet. . The device may include, for example, one or more of a processor 410, an input device 420, an output device 430, a storage device 440, and a communication device 460. The input device 420 and the output device 430 may generally correspond to those described above, and may be connectable or integrated with a computer.

Input device 420 may be any suitable device that provides input, such as a touch screen, keyboard or keypad, mouse, or voice-recognition device. The output device 430 can be any suitable device that provides output, such as a touch screen, a haptic device, or a speaker.

Storage device 440 may be any suitable device that provides storage such as electrical, magnetic or optical memory, including RAM, cache, hard drive, or removable storage disk. Communication device 460 may include any suitable device capable of sending and receiving signals over a network, such as a network interface chip or device. The components of the computer may be connected wirelessly or via a physical bus, for example, in any suitable manner.

Software 450 that may be stored in storage device 440 and may be executed by processor 410 is, for example, programming (e.g., implemented in a device as described above) that implements the functionality of the present disclosure. ) Can be included.

The software 450 is also used by or in connection with an instruction execution system, equipment or device as described above, capable of fetching and executing instructions associated with software from the instruction execution system, equipment or device. May be stored and/or transmitted within any non-transitory computer-readable storage medium. In the context of the present disclosure, a computer-readable storage medium can be any medium, such as storage device 440 that can contain or store programming for use by or in connection with an instruction execution system, equipment or device. .

The software 450 is also used by or in connection with an instruction execution system, equipment or device as described above, capable of fetching and executing instructions associated with software from the instruction execution system, equipment or device. To be propagated within any transmission medium. In the context of the present disclosure, a transmission medium may be any medium capable of communicating, propagating or transmitting programming for use by or in connection with an instruction execution system, equipment or apparatus. Transmission readable media may include, but are not limited to, electronic, magnetic, optical, electromagnetic or infrared wired or wireless propagation media.

Device 400 may be connected to a network, which may be any suitable type of interconnected communication system. The network can implement any suitable communication protocol and can be protected by any suitable security protocol. The network may include any suitable arrangement of network links capable of implementing the transmission and reception of network signals such as wireless network connections, T1 or T3 lines, cable networks, DSL or telephone lines.

Device 400 may implement any operating system suitable for operating on a network. The software 450 may be written in any suitable programming language such as C, C++, Java or Python. In various embodiments, application software implementing the functionality of the present disclosure may be deployed in other configurations through a web browser, for example in a client/server arrangement or as a web-based application or web service.

While the present disclosure and examples have been fully described with reference to the accompanying drawings, it should be noted that various changes and modifications will be apparent to those skilled in the art. It is to be understood that such changes and modifications are included within the scope of the disclosure and examples as defined by the claims.

For purposes of explanation, the foregoing detailed description has been described with reference to specific embodiments. However, the above exemplary discussion is not intended to be exhaustive or to limit the invention to the precise form disclosed. Many variations and modifications are possible in view of the above teaching. The embodiments have been selected and described to best explain the principles and practical applications of the technology. Those skilled in the art can best utilize the technology and various embodiments, along with various modifications suitable for the particular use contemplated by this.

SEQUENCE LISTING <110> Nanjingjinsirui Science & Technology Biology Corp. <120> CODON OPTIMIZATION <130> 75989-20004.40 <140> Not Yet Assigned <141> Concurrently Herewith <160> 8 <170> FastSEQ for Windows Version 4.0 <210> 1 <211> 5 <212> RNA <213> Artificial Sequence <220> <223> Synthetic Construct <400> 1 auuua 5 <210> 2 <211> 13 <212> DNA <213> Artificial Sequence <220> <223> Synthetic Construct <400> 2 tgyygatgyy yyy 13 <210> 3 <211> 6 <212> DNA <213> Artificial Sequence <220> <223> Synthetic Construct <400> 3 aggagg 6 <210> 4 <211> 10 <212> DNA <213> Artificial Sequence <220> <223> Synthetic Construct <400> 4 accaccatgg 10 <210> 5 <211> 10 <212> DNA <213> Artificial Sequence <220> <223> Synthetic Construct <400> 5 gccaccatgg 10 <210> 6 <211> 29 <212> DNA <213> Artificial Sequence <220> <223> Synthetic Construct <400> 6 tctctttaca ttttggtctc tatactaca 29 <210> 7 <211> 738 <212> DNA <213> Artificial Sequence <220> <223> Synthetic Construct <400> 7 atgtctaagg gagaagagct gtttaccggc gtggtcccta tcctggtgga gctggacggc 60 gatgtgaacg gccagaaatt cagcgtgtcc ggcgagggcg aaggcgacgc cacctacggc 120 aagctgacac tgaagttcat ctgcaccacc ggcaagctgc ctgtcccttg gccaacactg 180 gtgaccacct tcagctacgg agtgcaatgc ttctccagat accctgacca catgaagcag 240 cacgatttct ttaaatctgc catgcccgag ggctacgtgc aggaacggac catcttctac 300 aaggacgacg gaaattacaa gaccagagcc gaggtgaagt tcgagggcga caccctggtg 360 aaccggatcg agctgaaggg catcgacttc aaagaggatg gcaacatcct gggccacaag 420 atggaataca actacaactc ccacaacgtg tacatcatgg ccgacaagcc taagaacggc 480 atcaaggtga acttcaagat cagacacaac atcaaggacg gctctgtgca gctggccgac 540 cactaccagc agaacacccc catcggcgac ggccctgtgc tgctgcctga taaccactat 600 ctgtctacac agtccgctct gtccaaagat cctaacgaga agcgggacca catgatcctg 660 ctggagttcg tgaccgccgc tggcatcacc catggcatgg acgagctgta caaggactac 720 aaggacgatg acgacaag 738 <210> 8 <211> 1290 <212> DNA <213> Artificial Sequence <220> <223> Synthetic Construct <400> 8 atgtccctgc actttctgta ctactgctct gagcctaccc tggacgtgaa gatcgccttc 60 tgtcagggct tcgataagca ggtggacgtc tcctatatcg ctaagcacta caacatgagc 120 aaatccaagg tggacaacca gttctactct gtcgaggtgg gcgactctac cttcaccgtg 180 ctgaagagat accagaacct gaaacccatc ggctccggcg ctcagggcat cgtgtgcgcc 240 gcttacgacg ccgtgctgga tagaaacgtg gccatcaaga agctgagccg gcctttccag 300 aaccagacac acgctaagcg ggcctacaga gagctggtcc tgatgaagtg cgtgaaccac 360 aagaacatca tctccctgct gaatgtgttc acccctcaga aaaccctgga agagttccag 420 gatgtgtacc tggtgatgga actgatggac gccaacctgt gccaggtgat ccagatggaa 480 ctggaccacg agcggatgtc ctacctgctg taccagatgc tgtgtggcat caagcacttg 540 catagcgctg gcatcatcca cagagatctg aaaccttcta acatcgtggt gaagtccgac 600 tgcaccctga agatcctgga cttcggcctg gccagaaccg ctggcacctc tttcatgatg 660 acaccctacg tggtgaccag atactaccgg gcccctgaag tgatcctggg catgggctac 720 aaggagaacg tggacatctg gtccgtggga tgcatcatgg gcgagatggt cagacacaag 780 atcctgttcc ccggaagaga ttacatcgac cagtggaaca aggtgatcga gcagctgggc 840 accccttgtc ctgagttcat gaagaaactg cagcctaccg tgcggaacta cgtggaaaac 900 cggcctaagt acgccggcct gacctttcca aagctgttcc ctgactctct gttccccgct 960 gacagcgagc acaacaagct gaaagcctct caggccagag atctgctgtc caagatgctg 1020 gtgatcgacc ctgctaagag aatctccgtg gacgatgccc tgcagcaccc ctacatcaac 1080 gtgtggtacg accctgctga ggtggaagcc cctccacctc agatctacga caagcagctg 1140 gacgaaagag agcacaccat cgaggagtgg aaggagctga tctataaaga agtgatgaac 1200 tccgaggaaa agaccaagaa cggcgtggtc aagggccagc cttccccctc tgctcaggtg 1260 cagcaagact acaaggacga tgatgacaag 1290

Claims (42)

A computer-implemented method for optimizing nucleic acid sequences for protein expression in a host, the method comprising:
a) receiving an initial population set, wherein the initial population set comprises a plurality of initial candidate nucleic acid sequences capable of expressing a protein; And
b) Based on the initial population set, computer-assisted NSGA-III algorithms or variations thereof are used to perform optimization of the harmonic index, codon context index, and outlier index, thereby expressing the protein. Obtaining a plurality of optimized nucleic acid sequences,
Here, the harmony index of the candidate nucleic acid sequence represents the consistency of the frequency distribution of synonym codons between the plurality of highly expressed genes and the candidate nucleic acid sequence,
Wherein the codon context index of the candidate nucleic acid sequence is a measure for placing synonym codons in appropriate positions, and
Wherein the outlier index of the candidate nucleic acid sequence is a measure of the negative effect of a plurality of predetermined sequence features on the candidate nucleic acid sequence. The method of claim 1, further comprising providing an output representative of at least one optimized nucleic acid sequence of the plurality of optimized nucleic acid sequences. The method of claim 1 or 2, wherein receiving the initial population set comprises:
Receiving a protein sequence;
Generating an initial population set based on the received protein sequence. The method of claim 1 or 2, wherein receiving the initial population set comprises:
Receiving a nucleic acid sequence;
Translating the received nucleic acid sequence into a protein sequence;
Generating an initial population set based on the protein sequence. The method of any one of claims 1 to 4, wherein the initial population set is of a predetermined size. 6. The method of any one of claims 1-5, wherein the initial population set comprises a binary representation of a plurality of initial candidate nucleic acid sequences. The method according to any one of claims 1 to 6, wherein the step of performing optimization of the harmonic index, codon context index and outlier index comprises:
Maximizing the harmonic index;
Maximizing the codon context index; And
Minimizing the outlier index. The method of any one of claims 1 to 7, wherein the step of performing optimization of the harmonic index, codon context index and outlier index comprises:
For each initial candidate nucleic acid sequence in the initial population set, calculating each harmony index value, each codon context index value, and each outlier index value for each initial candidate nucleic acid sequence;
Allocating a plurality of suitability values corresponding to a plurality of initial candidate nucleic acid sequences based on the calculation;
Sorting a plurality of initial candidate nucleic acid sequences based on the plurality of suitability values; And
Incorporating a subset of the sorted plurality of initial candidate nucleic acid sequences into a subsequent population set. The method of claim 8, further comprising the following steps:
Generating an offspring population based on the initial population; And
Including the progeny population in a subsequent population set. The method of claim 9, wherein the progeny population is generated through binary tournament selection, cross/recombination, mutation, or any combination thereof. 11. The method of any one of claims 8-10, wherein the initial population set and the subsequent population set are of the same size. The method according to any one of claims 1 to 11,
The step of performing optimization of the harmonic index, codon context index and outlier index includes a plurality of iterations,
The i-th iteration of the plurality of iterations is:
receiving a population set of nucleic acid sequences corresponding to the (i-1) th iteration;
associating each nucleic acid sequence of the population set corresponding to the (i-1) th iteration with a non-domination level;
Classifying nucleic acid sequences in the population set corresponding to the (i-1) th iteration based on the associated non-dominant level;
generating a population set corresponding to the i-th iteration, wherein the population set corresponding to the i-th iteration is a subset of the classified nucleic acid sequence corresponding to the (i-1)th iteration and the (i-1)th iteration Comprising a progeny population generated based on the classified nucleic acid sequence corresponding to the repeat; And
Determining whether to proceed to the (i+1) th iteration using the population set corresponding to the i-th iteration based on the one or more termination conditions. The method of claim 12, wherein the step of associating each nucleic acid sequence with a non-dominant level comprises, for each nucleic acid sequence of the population set corresponding to the (i-1) th iteration, each harmony index value, each codon context index value, and Calculating each outlier exponent value. The method of claim 10 or 11, wherein generating a population set corresponding to the i-th iteration comprises:
associating at least one nucleic acid sequence of the classified nucleic acid sequences corresponding to the (i-1) th iteration with one of the plurality of predetermined reference points. The method according to any one of claims 10 to 12, wherein the one or more termination conditions reach a fixed number of iterations and reach the best fit and do not produce better results, the minimum criterion of a solution that is close to optimal is in some solutions. Satisfied by, or any combination thereof. The method of any one of claims 1 to 15, wherein the harmony index of the candidate nucleic acid sequence is calculated based on the formula: H = 1-D( F _hs ,F _{ts ),}
In the above, D() represents a distance function;
F _hs comprises a vector comprising the frequency of synonym codons of a plurality of amino acids in a plurality of highly expressed genes; And
F _ts comprises a vector comprising a frequency of synonym codons of a plurality of amino acids within the coding gene of the candidate nucleic acid sequence. The method of claim 16, wherein D() represents a function measuring the distance between two vectors. The method of claim 17, wherein D() is a distance function including, but not limited to, a Euclidean distance, a cosine distance, a Manhattan distance, or a Minkowski distance of two vectors. The method of claim 18, wherein the frequency of synonym codons in a plurality of highly expressed genes or candidate nucleic acid sequences is defined as follows:

The method of any one of claims 1 to 19, wherein the codon context index of the candidate nucleic acid sequence is calculated based on the formula: CC = 1-D( F _hcc ,F _{tcc ),}
In the above equation, D() represents a distance function;
F _hcc comprises a vector comprising the frequency of synonym codon pairs of two consecutive amino acids in a plurality of highly expressed genes; And
F _tcc comprises a vector comprising the frequency of synonym codon pairs of two consecutive amino acids within the coding gene of the candidate nucleic acid sequence. The method of claim 20, wherein D() represents a function measuring the distance between two vectors. The method of claim 21, wherein D() is a distance function including, but not limited to, a Euclidean distance, a cosine distance, a Manhattan distance, or a Minkowski distance of two vectors. The method of any one of claims 20-22, wherein the frequencies of synonym codon pairs of a plurality of highly expressed genes or candidate nucleic acid sequences are defined as follows:

The method of any one of claims 1 to 23, wherein the outlier index is an equation

Is calculated based on
Wherein N is the number of a plurality of predetermined sequence features;
fi(x) represents the penalty scoring function of the i-th sequence feature among a plurality of predetermined sequence features; And
wi denotes a relative weight associated with fi(x). The method of claim 24, wherein the plurality of predetermined features comprises:
GC-content value,
CIS element,
Repeating Element,
RNA splicing site,
Ribosome binding sequence,
the minimum free energy of the mRNA, or
Any combination of these. 25. The method of claim 24, wherein the plurality of predetermined features are identified based on a selected expression system. 27. The method of any one of claims 1-26, wherein the modification of the NSGA-III algorithm comprises an EliteNSGA-III algorithm or an NSGA-II based immune algorithm. 28. The method of any one of claims 1-27, wherein the step of performing optimization of the harmonic index, codon context index and outlier index comprises:
Ranking the plurality of optimized nucleic acid sequences by descending order of harmonic index, then descending order of codon context index, and then ascending order of outlier index;
Selecting one or more top-level optimized nucleic acid sequences for synthesis. The method of any one of claims 1-28, further comprising:
c) removing a predetermined adverse subsequence or motif from the optimized nucleic acid sequence of the plurality of optimized nucleic acid sequences. 30. The method of claim 29, wherein the predetermined adverse subsequence or motif is identified based on analysis of a plurality of text portions. The method of claim 29, wherein removing the predetermined adverse subsequence or motif comprises:
Identifying the predetermined adverse subsequence or motif in the optimized nucleic acid sequence;
Identifying a plurality of synonym codons based on the identified predetermined adverse subsequence or motif;
Selecting a synonym codon from the plurality of synonym codons for substitution with the identified predetermined adverse subsequence in the optimized nucleic acid sequence. The method of any one of claims 1-31, wherein at least one of the harmonic index, the codon context index, and the outlier index is calculated based on one or more characteristics of a plurality of highly-expressed genes from one or more databases. Way. 33. The method of claim 32, wherein the one or more characteristics comprise codon frequency, synonym codon frequency, codon pair frequency, or a combination thereof. 34. The method of any one of claims 1-33, further comprising setting one or more parameters, wherein the one or more parameters are the size of the population set, the number of divisions, the distribution index for the simulated binary intersection, A crossover ratio for a simulated binary crossover, a mutation ratio for a bit flip mutation, a distribution index for a bit flip mutation, or any combination thereof. A non-transitory computer-readable storage medium storing one or more programs, wherein the one or more programs cause the electronic device to perform the method of any one of claims 1 to 34 when executed by one or more processors of the electronic device. A storage medium containing instructions to do. A system for optimizing a nucleic acid sequence for protein expression in a host, the system comprising:
One or more processors;
Memory; And
One or more programs, wherein the one or more programs are stored in the memory and are configured to be executed by the one or more processors, and the one or more programs include instructions for performing the method of any one of claims 1 to 34. Containing, system. An electronic device for optimizing a nucleic acid sequence for protein expression in a host, the device comprising means for performing the method of claim 1. A program product stored in a recordable medium for optimizing a nucleic acid sequence for protein expression in a host, the program product comprising computer software for performing the method of claim 1. An isolated nucleic acid molecule comprising an optimized nucleic acid sequence obtained from the method of any one of claims 1 to 34. A vector comprising the isolated nucleic acid molecule of claim 39. A recombinant host cell comprising the isolated nucleic acid molecule of claim 39 or the vector of claim 40. A method of expressing a protein in a host cell, the method comprising:
(a) obtaining an optimized nucleic acid sequence for protein expression in a host cell using the method of any one of claims 1 to 34;
(b) synthesizing a nucleic acid molecule comprising the optimized nucleic acid sequence;
(c) introducing the nucleic acid molecule into the host cell to obtain a recombinant host cell; And
(d) culturing the recombinant host cell under conditions that permit expression of the protein from the optimized nucleic acid sequence.

Images